PROJECT SUMMARY Most studies of human disease have focused on peripheral blood which is easily obtained and amenable to current high-throughput technologies. However, important pathogenic mechanisms for many, if not most, human diseases arise and are propagated in tissue. We have taken two innovative approaches to understand in situ adaptive immunity in human inflammatory renal diseases. In the first, we have sorted B cells from renal biopsies of acute mixed allograft rejection (AMR) patients, subjected these cells to single cell (sc) RNA-Seq and then cloned the immunoglobulin variable regions from these same cells and expressed corresponding antibodies. In this way, we have been able to pair transcriptional state in single B cells with the antigenic specificity of the antibody they secrete. In the second, we have developed a novel approach to characterize the spatial architecture between different cell populations in human tissue and identify functional relationships. We did this by implementing a deep convolutional neural network (DCNN) that accurately identified both the position and shape of complex cells in tissue multicolor confocal micrographs. The DCNN output was then analyzed with a tuned convolutional neural network (TNN) to identify distance and T cell shape features that best distinguished between different T cell populations relative to DCs. We refer to this analysis pipeline as CDM3. In mice, CDM3 discriminated between cognate and non- cognate T cell interactions with DCs with a sensitivity and specificity similar to most two photon microscopy measures. In human lupus tubulointerstitial inflammation (TII), CDM3 both confirmed that myeloid DCs present antigen to CD4+ T cells in situ and identified infiltrating plasmacytoid DCs as an important antigen presenting cell in severe lupus TII. We propose to apply CDM3, scRNA-Seq and immunoglobulin repertoire technologies to dissect the in situ adaptive cell networks in human autoimmunity. Our overall hypothesis is that in many autoimmune and inflammatory diseases, self-propagating in situ adaptive cell networks drive local inflammation and tissue damage. This hypothesis will be tested in these Specific Aims: Aim 1: To determine mechanisms of in situ B cell selection in renal inflammation. Aim 2. Determine if differences in inflammation architecture underlie autoimmune disease heterogeneity.